- 67 Views
- Uploaded on
- Presentation posted in: General

Measurement of gas gain fluctuations

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Measurement of gas gain fluctuations

M. Chefdeville, LAPP, Annecy

TPC Jamboree, Orsay, 12/05/2009

- Introduction
- Motivations, questions and tools

- Measurements
- Energyresolution & electron collection efficiency
Micromegas-likemeshreadout

- Single electrondetectionefficiency and gas gain
TimePixreadout

- Energyresolution & electron collection efficiency
- Conclusion

- Final avalanche size obeys a probability distribution
Signal fluctuations impact on detector performance

- Spatial resolution in a TPC
- Energyresolution in amplification-basedgas detectors
- Minimum gain and ion backflow
- Detection of single electronswith a pixel chip

How doesitvarywithgas, field, geometry…?

gas gain G and parameterm

- Works wellwith Micromegas/PPC/MCP/single GEM
- With GEM stacks, distribution is more exponential

Micromegas, NIMA 461 (2001) 84

= b, relative gain variance

Micromegas

T. Zerguerras et al.

to be published in NIMA

GEM, Bellazzini, IEEE 06, SanDiego

1 mm gap PPC

NIMA 433 (1999) 513

3 GEMs, NIMA 443 (2000) 164

PPC, Z. Phys. 151 (1958) 563

Micromegas

MCP+Micromegas, NIMA 535 (2004) 334

- Simulation, since recently within GARFIELD:
- Simulation of e- avalanche according to
MAGBOLTZ cross-section database:

study of gas & field

- Simulation of e- tracking at microscopic scale
in field maps (3D):

study of geometry

- Simulation of e- avalanche according to
- On the experimental side:
- Direct measurement of the distribution:
- High gains, low noise electronics, single electron source

- Indirect measurements
- Do not provide the shape but some moments (variance)
- Assuming Polya-like fluctuations, one obtains the shape

- Direct measurement of the distribution:
- In this talk, only indirect methods are presented
The Polya parameter m is deduced from:

- Trend of energy resolution and collection efficiency
- Trend of single electron detection efficiency and gas gain

- Energy resolution R and electron collection efficiency η
- R decreases with the efficiency according to
- R2 = F/N + b/ηN + (1-η)/ηN
- R2 = p0 + p1/η
p0 = (F-1)/N

p1 = (b+1)/N

- Measure R(η) at e.g. 5.9 keV, fix F and N, adjust b (i.e. m) on data

- Single electron detection efficiency κ and gas gain G
- κ increases with G as more avalanches end up above the detection threshold t
Integral can be calculated for integer value of m

- Count the number of e- from 55Fe conversions (N) with TimePix
- Measure N(G), adjust κ(G,m) on this trend, keep m for which the fit is best

- κ increases with G as more avalanches end up above the detection threshold t

- Measure 1: R(η)
- InGrid on bare wafer
- Preamp/shaper/ADC
- 55Fe 5.9 keV X-ray source
- Ar-based gas mixtures
with iC4H10 and CO2

- Measure 2: κ(G)
- InGrid on TimePix chip
- Pixelman and ROOT
- 55Fe 5.9 keV X-ray source
- Enough diffusion for counting
- 10 cm drift gap
- Ar 5% iC4H10

- Vary the collection efficiency with the field ratio
- Record 55Fe spectra at various field ratios
- Look at peak position VS field ratio
define arbitrarily peak maximum as η = 1

- Look at resolution VS collection
- Adjust b on data points

- Look at peak position VS field ratio

- Fit function:
- R = √(p0 + p1/η)
- p0 = (F-1)/N
- p1 = (b+1)/N
- F = 0.2 in all mix.
- N = 230 in Ar/iso
- N = 220 in Ar/CO2

- Record 55Fe spectra at various field ratios
- Look at peak position VS field ratio
define arbitrarily peak maximum as η = 1

- Look at resolution VS collection
- Fix F and N, adjust b on data points

- Look at peak position VS field ratio

- Rather low Polya parameter 0.6-1.4
- May be due to a poor grid quality
- Curves do not fit very well points
- Could let F or/and N free

(TimePix) threshold

- Trend depends on:
- the threshold t, the gas gain G and m

55Fe X-ray E0: 5.9 and 6.5 keV

Photo-e-

Auger-e-

TimePix chip

- Considering55Fe conversions:
the efficiencyisproportionnal to the number of detectedelectronsat the chip

- Count the number of electronsatvarious gains
- In the escape peak!
- Applycuts on the X and Y r.m.s. of the hits

55Fe cloud in Ar5iso after 10 cm drift

- Number of detected electrons at given voltage determined by
- Adjusting 2 gaussians on escape peak
- Kbeta parameters constrained by Kalpha ones
- 3 free parameters

- Number of detected electrons and voltage
- Use common gain parametrization
- Fix p2 (slope of the gain curve)
- 2 free parameters: t/A and ηN

Best fit for m = 3

Yields √ b =1/√m ~ 58 %

m rms (%) Chi2

1 100 39.43

2 71 3.08

3 58 1.16

4 50 1.41

5 45 1.60

- Also, ηN = 115 e-
- Upper limit on W(Ar5iso) < 25 eV
- Correcting for un-efficiency:
- Upper limit on F(Ar5iso) < 0.3

- Two methods to investigate gas gain variance
- Assuming Polya fluctuation, shape available for detector simulation
- Energy resolution and collection efficiency
simple (mesh readout)

but a certain number of primary e- and Fano factor have to be assumed

- Single e- detection efficiency and gain
powerful (provide not only m but W and F)

but a InGrid-equipped pixel chip is needed

- Another one not presented
- Energy resolution and number
of primary electrons

- Energy resolution and number

4 main lines in an 55Fe spectrum

- R2 = p0/N
- p0 = F+b